The present invention relates to a nuclear magnetic resonance spectrometer suitable for analyzing, in a liquid-solution, the structure and interaction of protein and organic molecules such as substrate and ligand interacting with the protein.
A method for analyzing organic matter utilizing nuclear magnetic resonance (NMR) has been making rapid progress in recent years. In particular, a method has been used in combination with the technique of strong superconductive magnets to make it possible to highly efficiently analyze the structure of an organic compound such as protein having a complicated molecular structure on atomic level. The present invention is directed to a nuclear magnetic resonance (NMR) spectrometer necessary for analyzing the structure on atomic level and interaction of protein molecules in an aqueous solution dissolving a small quantity of protein, especially, an energy spectrometer differing from a medical MRI diagnostic apparatus aiming at imaging of a tomogram of human body needing so-called millimeter class image resolution in that performance such as the magnetic field intensity being one order or more higher, the magnetic field uniformity being of four order and the stability being three order higher is needed, thus requiring quite different design technique and apparatus manufacture technique. A detailed description of a conventional high-resolution NMR spectrometer is given in “NMR for Protein” by Yoji Arata, published by Kyoritu-shuppan, pp. 33-54, 1996. As up-to-date inventions concerning the typical apparatus construction when the NMR is utilized for analysis of protein, one may refer to an invention relating to a superconductive magnet that shows the typical construction of a multilayer air-core solenoid coil as disclosed in JP-A-2000-147082, an invention relating to a signal detection technique that shows a cage type superconductive detection coil as disclosed in U.S. Pat. No. 6,121,776 and examples of signal detection technique based on a conventional barrel type or cage type coil as disclosed in JP-A-2000-266830 and JP-A-6-237912. According to these reports, all of the conventional high-sensitivity NMR spectrometers for protein analysis use a superconductive magnet unit constructed of solenoid coils used in combination to generate a magnetic field in the vertical directions, an electromagnetic wave at 400 to 900 MHz is irradiated on a sample and a resonance wave generated from the sample is detected by utilizing the barrel or cage type detection coil. Also, as shown in an example of U.S. Pat. No. 6,121,776, a contrivance is made to improve the S/N sensitivity ratio by utilizing a detector cooled to low temperatures with a view to decreasing heat noise during reception.
Historically, in the high-sensitivity NMR spectrometer, improvements in sensitivity have been achieved by a method in which the basic constituents of a system such as antenna and magnet are kept to remain unchanged but the center magnetic field intensity of the superconductive magnet is increased. Accordingly, the maximum NMR measurement sensitivity reported till now can be obtained with a NMR spectrometer at 900 MHz utilizing a large superconductive magnet having a center magnetic field of 21.1 tesla and the basic constituents of the spectometer remain unchanged as compared to those in the prior art of JP-A-2000-147082. In the analysis of protein using a liquid-solution, the improved center magnetic field is effective to clarify separation of improvement in sensitivity and chemical shift.
In attaining the effect of improved sensitivity attributable to the form or shape of detection coil, it has been known, for example, as described in “Book of NMR” by Yoji Arata, published by Maruzen, PP. 325-327, 2000, that the solenoid coil conventionally used as detection coil is advantageous over the barrel or cage type in various points. For example, the solenoid coil is advantageous in easy controllability of impedance, filling factor and efficiency of RF magnetic field. But, according to the literature, when the sensitivity is thought much of in such an application of measuring protein dissolved by a small quantity in an aqueous solution, winding a solenoid coil around a sample tube placed vertically to the magnetic field is practically impossible and in general, is not utilized. In an exceptional application where highly sensitive measurement is carried out by using a small quantity of sample solution, the above technique is limitedly utilized through a method utilizing a particularly designed micro-sample tube to carry out measurement by using a special probe.
Recently, in a special example disclosed in JP-A-11-248810, a method is contrived according to which a bulky magnet of high-temperature superconductivity is magnetized in the horizontal direction and a NMR signal is detected with a solenoid coil. Further, JP-A-7-240310 discloses a method for constructing a superconductive magnet and a cooling container to meet a general NMR application directed to eliminate constraints on the top or ceiling height of apparatus. However, any method of improving the detection sensitivity necessary for analyzing protein and any method of technically coping with magnetic field uniformity and temporal stability of magnetic field have not been known yet.
Recently, with needs for study of protein promoted, needs for analyzing a sample in which protein has a small degree of solubility to water have increased and there is a need of improving the sensitivity of measurement of NMR. To adapt the NMR spectrometer to the needs as above, the measurement sensitivity must be improved while maintaining a sample space comparable to that of the conventional apparatus and besides the maintenance of the stability of a superconductive magnetic field over a long time of data integration is indispensable. The improved measurement sensitivity is particularly advantageous in that for samples having substantially the same solubility, not only the measurement time can be shortened but also the sampling amount can be decreased, thereby ensuring that the protein of small solubility can be analyzed to advantage. Accordingly, the NMR spectrometer used for analysis of protein is required of far more excellent detection sensitivity and stability than those in the conventional NMR and in addition, is required to have ability to detect NMR signals accurately and stably over a long time of one week or more. This is because if the magnetic field varies during measurement, the peak of NMR signal is caused to shift and especially, in measurement of interaction, the peak shift due to interaction cannot be discriminated from that due to instability of the magnetic field. If the magnetic field is non-uniform, desired peaks overlap each other, raising a problem that discrimination of interaction is difficult to achieve. Therefore, it should be noticed that, in future NMR techniques aiming at performing various kinds of analysis of protein, development of new technology not lying on mere extension of the conventional general NMR spectrometers will be needed.
For example, specification of magnetic field uniformity in the general NMR spectrometer is 0.01 ppm in a sample space, that is, 0.01 ppm in terms of temporal stability. When this value is indicated in terms of proton NMR for general 600 MHz use, a permissible error of 6 Hz results. In the case of the aforementioned analysis of interaction of protein, however, spatial and temporal resolution of at least 1.0 Hz or less is required and preferably, 0.5 Hz or less is needed. In a method capable of implementing the magnetic field stability and the temporal stability of magnetic field, the construction of superconductive magnet and detection coil must be optimized. Accordingly, the performance of the conventional, generally used NMR spectrometer is insufficient and the stability and magnetic field uniformity higher by one order or more than those of the conventional spectrometer are required.
In the prior arts, the sensitivity is managed to be improved by relying on improvements in magnetic field intensity and as a result, the apparatus is increased in size and to cope with problems of leakage magnetic field and floor strength, there arises a new problem of installation capability such as needs for a dedicated building. Further, disadvantageously, the cost of a superconductive magnet increases. The improved sensitivity has an upper limit of about 21 T because of constraints due to a critical magnetic field of a superconductive material and for more upgraded improvements in sensitivity, the advent of a technique for improving detection sensitivity based on a new means without resort to the magnetic field intensity has been desired.
The aforementioned high-sensitivity measuring method utilizing the solenoid coil can be used with a special sample tube for a very small quantity of sample and a special detection probe but it cannot be applied to analysis based on a general protein solution of about 10 cc. The method for generating a magnetic field in the horizontal direction by means of a strong magnet and detecting NMR signals by means of a solenoid coil as described in the example of JP-A-11-248810 can generate only a magnetic field of not greater than 10 T at the surface of a high-temperature superconductor, with the result that the magnetic field at a sample part is about several tesla at the most, thus proving that the method of interest cannot generate a magnetic field of 11 tesla or more necessary for analysis of protein, preferably, a magnetic field of 14.1 tesla or more in a desired sample space. Further, in this method, owing to the effect of a magnetic flux creep phenomenon of the high-temperature superconductor, the temporal stability 1.0 Hz/hour or less necessary for analysis of protein is substantially difficult to achieve. As regards the magnetic uniformity necessary for analysis of protein, non-homogeneity attributable to the manufacture process of a high-temperature superconductive bulky material also makes it difficult to attain the magnetic field uniformity within 1.0 Hz in terms of proton NMR frequency in a space defined by 10 mm diameter×20 mm length.
As described above, while a breakthrough technique meeting the needs for analysis of protein is desired to be developed in connection with the conventional techniques, the advent of a new solving method for further improvements in sensitivity has been desired under the present-day circumstances that improving the sensitivity based on the magnetic field has reached limits.
For the purpose of conducting an efficient and accurate. Analysis of the interaction of protein in a liquid-solution with low molecules such as substrate and ligand, for which needs are considered to increase in future, it is empirically preferable that a suitable quantity of sample be measured at 600 to 900 MHz and with a center magnetic field of about 14 to 21 T and the measurement sensitivity be increased beyond the present one to increase the throughput. Generally, in a spectrometer operating at 800 MHz or more, for the purpose of making full use of the superconductive characteristics to an extreme, operation is carried out by depressurizing liquid helium at 4.2 K and excessively cooling it to 1.8 K. Therefore, complexities in apparatus operation are aggravated and maintenance is laborious. In addition, the magnetic unit Increases in size to increase leakage magnetic field and typically, a dedicated building is needed. Especially, the leakage magnetic field in the vertical direction increases as the center magnetic field increases in the conventional system, so that in an apparatus of 900 MHz class, for instance, a leakage magnetic field occurs extending up to 5 m in the height direction, and from the viewpoint of apparatus installation, there needs a tall building of high ceiling. As a result, the construction cost increases disadvantageously. Further, the conventional 900 MHz superconductive magnet is sized such that only a magnet part has a diameter of 1.86 m and a height of several meters, as described in IEEE. Transactions on Applied Superconductivity, Vol. 10, No. 1, page 728-731.
The present invention intends to provide a novel NMR spectrometer in which the measurement sensitivity of NMR signals can be increased hit at least 2.5 times or more of that in the conventional apparatus at about 600 MHz (14.1 T) under a condition that a normal sample tube of 5 to 10 mm diameter is mainly used and a sample liquid-solution is charged in the tube up to a height of about 30 mm and the temporal stability and spatial uniformity of a superconductive magnet necessary for analysis of protein can be provided. In the construction of the present invention, the operating temperature of a system is not set to 4.2 K. By applying the present invention, it is also possible to aim at achieving extremity performance but depending on applications, operation at the conventional magnetic field limit 21.1 T that is, at 900 MHz and at 1.8 K can proceed and in that case, the sensitivity can be improved by 40% of that in the conventional system, proving that overcoming the detection sensitivity limit attributable to magnetic field intensity, conventionally unattainable, can succeed for the first time.